The long-term objective of this proposal is to develop a random functional screening system which might be used in the future by researchers to search for new tumor suppressor genes for treatment of human cancer. By combining a novel retrovirus packaging method with a modified inducible expression system, we have recently constructed a novel human complementary DNA (cDNA) library. The cDNA library consists of 5 x 105 independent clones, which contain random cDNA insert(s) derived from human normal fibroblasts. Of importance, in this cDNA library, expression of the insertional genes can be reversibly turned on and off. Preliminary screening of the cDNA library under permissive conditions has revealed at least 34 clones containing insertional genes whose expression conferred tumor cells lower growth rates and/or reduced metastatic potential. The major emphasis of this proposal will be placed on these genes. Candidate therapeutic genes for human cancer will be initially selected based on their capability to suppress tumor cell growth as measured in vitro by cell growth rate, soft agar colony formation and invasive potential in Matrigel, and in vivo by nude mouse tumorigenicity and tumor regression assays. Assuming all the criteria in the initial selection step have been satisfied, the candidate therapeutic gene(s) will then be used to treat human xenograft tumors in experimental animal models via plasmid/liposome complex or viral vector-mediated gene transfer. Meantime, the best candidate gene(s) for human cancer gene therapy selected in this system will be further studied at the basic science level. These include complete DNA sequencing, chromosomal mapping, measurement of deletion frequency of the gene(s), if applicable, in certain tumor types, and developing antibodies against the deduced proteins to facilitate elucidating the gene function. Preferably, the candidate gene products are small diffusible, or secretable proteins, therefore, the future therapeutic gene(s) for cancer might not have to be delivered to every tumor cell in a patient and systemic therapy might be feasible. Finally, the side effects as well as short- and long-term toxicity of the therapeutic gene expression in patients will be determined pre-clinically in normal human fibroblasts and in transgenic mouse models. Advanced techniques in molecular medicine, including vector construction, viral vector packaging, gene therapy in experimental animal models, transgenic mice and standard molecular and cellular biology methods will be used for these various studies. Fulfilling the goal of this proposal would result in identification of a series of novel therapeutic genes for human breast cancer and other cancers, with much of the pre-clinical efficacy and safety data already being accumulated which are likely required by RAC and FDA approvals of any Phase I clinical trials for human cancer gene therapy.